mysingingmonstersfandomcom-20200222-history
Talk:Glowbes
Colour wheel around island I wonder: suppose one wanted to create a colour wheel running around the perimeter of an island. What's the "resolution"? -- that is, how many 1x1 spaces are there in that perimeter? And what's the algorithm to create the correct colours by combining red, yellow, and blue Glowbes? "Merge a red and a yellow to give orange, 50% red and 50% yellow. Then merge that with another red, that's 75% red and 25% yellow. Then another yellow, that's 37.5% red and 62.5% yellow. Then merge that with a 50% red / 50% yellow..." BunsenH (talk) 06:55, April 26, 2015 (UTC) : I wouldn't dream of touching the algorithm part, but in terms of the perimeter of an island, it depends on what geometry is acceptable. I would suggest only the main part of the island (so ignore the pedestals), and at least ignore the extra grid cells. Depending on how one draws the convex corners, you get 91 cells (open corners) or 128 cells (connected corners). : Alternatively, if you get rid of some of the protrusions and make it completely symmetrical (so that the bottom of the island mirrors the top as well), there's 120 cells. : None of these would be particularly easy to mix, and figuring out the exact mixing formula to get to any of these I'll happily leave as an exercise for the reader. : Perhaps more viable for people is to go the other way around. What's the smallest easily obtained (mixing results with primary colors) color step between Glowbes that can be made to fit nicely on an island? Using the symmetrical option above as a constraint, that smallest step would be 3.125% of a color at each step (mixing e.g. red+yellow+y+y+y+y to get 3.215% of red), yielding 96 glowbes total. The question then becomes how to fit those 96 glowbes on the island. An oval shape, as depicted, works. DemEyes (talk) 16:08, August 3, 2017 (UTC) I wonder How do you create a grey Glowbe? Very sure? Qwertyxp2000 the second (talk) 00:55, April 27, 2015 (UTC) : You're trying to get proportions of ⅓R + ⅓Y + ⅓B starting with Glowbes that are initially just R or Y or B. The only mathematical operation you're allowed to perform is to average two values -- that is, to add them together and divide the result by 2. You'll never be able to get there exactly but you can get arbitrarily close, to the limit of the precision of the calculation. : This is getting into the fun territory of representing fractions in binary. :-) BunsenH (talk) 02:33, April 27, 2015 (UTC) : I'm quite willing to explain further if you wish, but I don't want to waste people's time (including my own) otherwise. BunsenH (talk) 04:25, April 27, 2015 (UTC) Friendly advice Someone on Facebook said this... Starting with red, and going clockwise, red+orange=red/orange, red+yellow=orange, yellow+orange=yellow/orange, (yellow), yellow+green=lime, blue+yellow=green, blue+green=aqua, (blue), blue+orange=dk.blue, blue+pink=purple, red+green=brown, yellow+pink=white, blue+red=pink, and red+pink=hot pink. These are the basic ones, but you can make even more. Hope this helps. After this pic, I fuzed aqua with blue to get a more turquoise color. There are a lot of possibilities. Qwertyxp2000 the second (talk) 08:43, April 27, 2015 (UTC) Red and blue Are you very sure Red and Blue create Violet with these Glowbes? Qwertyxp2000 the second (talk) 05:23, April 28, 2015 (UTC) : Yes, by experiment. What were you expecting? BunsenH (talk) 05:53, April 28, 2015 (UTC) ::: I hope this is the case. How to make pink? Qwertyxp2000 the second (talk) 08:59, April 28, 2015 (UTC) ::::: That's a bit hard to answer, since it depends partially on perception and circumstance. When one is mixing paints, pink is often just red mixed with white, sometimes with a bit of blue thrown in. With lights, you get something pinkish with a colour balance that has some white light and some purplish-red -- how that's achieved depends on the colour system. What's going on here in the game is weird... combining R/Y/B as light colours, and the lights don't actually illuminate their area, and you're judging the colour partly from the opaque "base" of the Glowbe, partly from the transparent "body" and "halo", and partly from the "light" inside which is opaque, brightened, and desaturated. ::::: I think you're looking for red + violet, which would give a pink / magenta / fuchsia. BunsenH (talk) 16:43, April 28, 2015 (UTC) ::::: I should point out that with regard to this subject, I'm a moderately-informed amateur. Anything that I write about how the Glowbe colours work is deduction based on observation, and may not be correct. I can't look at how the game actually does its calculations or draws things. I know for sure that I don't know everything about how the monsters are drawn. For example, I haven't figured out how transparency works for the monsters; when I alter the picture files, stuff that ought to be completely transparent is still visible. BunsenH (talk) 22:56, April 28, 2015 (UTC) Glowbes colour palette Just thought I'd add this here because I can't seem to add it to the slideshow thing Maybe can someone else try? I'm useless on here Hewy (: (talk) 10:14, May 10, 2015 (UTC) Super! By the way, tell me every combination possible to get such nice colours. Qwertyxp2000 the second (talk) 20:39, May 10, 2015 (UTC) To add to gallery, click on the gallery's "add photo" button, then click on "add a photo" once the window appears, then find your photo(s) necessary. Qwertyxp2000 the second (talk) 20:40, May 10, 2015 (UTC) Red and blue concern I am still concerned on whether you know your colours. Although red and blue make purple in real paints, the Glowbes somehow create the combination of pinkish when I fuse them together. Qwertyxp2000 the second (talk) 06:25, May 13, 2015 (UTC) Quaternary, quinary Given variations in display hardware, and in human colour perception, I think that when you're getting down to quaternary colours -- to say nothing of quinary! -- you're going to get a lot of argument about the proper names for the various shades. BunsenH (talk) 01:51, June 17, 2015 (UTC) Can't induce seizures, per neurologist I consulted the in-house neurologist. She says that the Glowbes are extremely unlikely to induce epileptic seizures. They don't blink quickly enough (their fastest blink rate is about 2.4 Hz, and the "best" rate to induce seizures is around 3 Hz), they don't occupy enough of the visual field, and there isn't enough contrast with the background. BunsenH (talk) 16:01, July 6, 2015 (UTC) : (And seriously, this is the kind of thing that game designers are careful about, now that the problem is known!) BunsenH (talk) 16:02, July 6, 2015 (UTC) An issue with particular colors It seems that I cannot usually find specific names of particular colors easily. It appears that some website on 9GAG says that males are not that likely to find the specific names of particular colors easily. And most of all wiki users are male. Qwertyxp2000 the second (talk) 08:00, July 8, 2015 (UTC) Unusual colour result / random factor? Having just recently started the game on iPad, I have been stockpiling the primary and secondary colours before making others. Strangely, one of my resulting Glowbes has come out dark blue, supposedly a quinary colour. While I can see my accidentally mixing a primary and a secondary to get a tertiary without meaning to, how could I possibly have gotten a quinary unless there is a random factor (like how you can get an Ethereal 1% of the time when combining a 4-element monster with a 3-element monster). GlassDeviant (talk) 13:35, November 11, 2015 (UTC) Subtractive color model? The article notes that the Glowbes appear to be using a subtractive color model. How certain are we that this is true? For that matter.. how certain are we on the base colors of the Glowbes? The reason I ask is because if blue is considered truly blue, and yellow as truly yellow, then in a subtractive color system, they would make black. Instead, we get green. That would be the result of cyan and yellow (just as with print inks). However, cyan and red, in a subtractive color model, make black. Instead, we get something that is similar to, but not quite, magenta. ( For child-friendliness, I understand the choice of colors... but boy do I wish it was RGB and additive. ) Just curious - it doesn't matter much for designs / rainbows / fades. DemEyes (talk) 13:53, August 3, 2017 (UTC) Old white Glowbe mixing strategy I'm replacing the section on mixing Glowbes to get 'White'. It is a very inefficient approach that gets less and less efficient with each step proposed. Specifically, to get to a white state that does not benefit from any further mixes, you need to use the Fuzer 205 times for conjuring / fuzing. (The original claim at step 12 would require 125 steps, but in reality this leaves the Glowbe with a very subtle blue tinge (rgb(119,119,121), requiring one more mix which nearly doubles the effort). Contrast this to the more basic alternating Fuzing of Orange+Blue (proposed in the Forum as well, I believe) that I'm replacing it with, which only requires the use of the Fuzer 29 times, and is far simpler to keep track of. However, I'm keeping it in this discussion as a piece of curiosity. Its original research was rather well done, and I don't want to take away from the work they - Regdar1 - put in; there's simply a more efficient approach. Technical background: To make sure that a Glowbe was actually grey, and not introducing any color of its own, its body (cups and gloves) cannot be measured - no matter how many mixes, its colors always seemed off from the theoretical in a distinct direction on the hue wheel. The Glowbe's glow is added to whatever is shown behind the Glowbe. Which meant that in order to find the true color of the Glowbe's color, it had to be placed in front of a background element that was a neutral grey itself with sufficient surface to consistently sample the color. Cold Island may seem like a good candidate, but it has a blue tinge. Mammott is slightly beige, Toob is bluish, Squeed Statue and path stones too green, and the Recording Studio and Fuzer itself are too yellow. The Dragon Tower however has two large patches of pure grey (rgb(91,91,91)) at its base - perfect for placing a Glowbe in front. The measured colors, with lightness reduced by 25%, are what I used in the table cell backgrounds. Once the grey color was reached, it also became possible to determine the Glowbe's inherent color; a soft beige (48° on a hue wheel) THE FOLLOWING IS LARGELY DEPRECATED INFORMATION INTENDED ONLY FOR DISCUSSION White Glowbes If you managed to create a Glowbe that was ⅓ red, ⅓ yellow, and ⅓ blue, you would end up with a Glowbe showing no color at all -- that is, grey/white. It's mathematically impossible to get exactly ⅓R + ⅓Y + ⅓B, but you can get arbitrarily close with sufficient effort. This chart shows a way of getting extremely close, using a lot of glowbes and fuzing. Note that standard computer monitors only use 8 bits to describe the colour levels of red, green, and blue. That's a precision of 1 part in 256, or about 0.4%. So you'd be as close to white (equal amounts of the components) as you could tell on a standard computer monitor by the end of step 12 on the chart. High-end consumer display equipment gives 10 bits of precision to each of red, green, and blue. That is, each component's brightness is precise to 1 part in 1024, or about 0.1%. So you'd be as close to equal amounts as most really good monitors can show by the end of step 17 on the chart. Professional display equipment uses still higher precision, but that's more for trying to show both extremely bright and extremely dark objects at the same time. It wouldn't help you to see tiny colour differences between objects at about the same brightness. That's getting into the limitations of human colour perception. If you were to continue the pattern of fuzing the latest result with the result of the previous step at step 55 you would achieve 33.3333333333333000% of RYB. This Glowbe's color would be almost completely indistinguishable from pure white by any measurement equipment... not to mention far better precision than the game itself uses. ---- DemEyes (talk) 23:37, September 21, 2018 (UTC)